Monday, April 25, 2011

Initial look at lessons learned from Fukushima

A review of what went wrong, why, and what should be done in the future

Guest Blog Post by: Akira T. Tokuhiro Ph.D *

Fukushima after 1Following a magnitude 9.0 earthquake and as high as ~14 meter tsunami, the Fukushima  Dai-ichi (D1) and Dai-ni (D2) Nuclear Power Plants (NPPs, Units 1-4[U1-4] at D1, U5-6-2 at D2i) experienced a series of multiple incidents caused by inadequate cool down of decay heat in both the reactor and in the co-located spent fuel pool (SFP).

The reactors at D1, U1-6 were constructed as part of a GE/Hitachi/Toshiba collaboration and began commercial operation, during 1971-1979; U1-5 are GE-BWR, Mark-I, U6 is a Mark-II. Two GE ABWRs are due to start construction in April 2012.

(Photo right shows damage from hydrogen explosions (4 bottom, 3 second from bottom, and 1 top)

Impact of loss of power

Although the Units at D1 and D2 automatically shutdown at the onset of the quake and with near immediate loss of off-site power, the back-up diesel generator operated (~30minutes) until the tsunami inflicted considerable (unknown) damage to auxiliary and back-up systems (most prominently the back-up diesel general and batteries).

This initiated the onset of lack of decay heat cooling. Additional aftershocks continued for about one-week. During initial week, March 11-18, there were up to three larger (likely H2 explosion) explosions, vapor/steam jets and fires that further stressed the RPV, the containment and (weather) confinement buildings.

Damage to primary containment?

One of the later explosions conceivably damaged the primary (coolant) containment and thus, water found in the adjacent basement of the turbine building pointed to high-levels of radiation including fission products. Additional large volumes of contaminated water were found in the U-shaped electrical conduit ‘trenches’ off of U1-3 and spreading into other areas such as beneath the reactor site.

Outline of lessons learned

nuclear_power_plant_control_roomThis paper outlines the initial list of lessons learned from the multiple sequence of events, some interpretations of the news releases and the aspects of safety culture that contrast Japan and the U.S. during crisis management.

It is based largely on events of the first three weeks and professional interpretation of publically accessible information. It is being released without peer review and in this summary form. Only the provisionally conclusive lessons learned are noted below.

1) Nuclear R&D institutions must consider alternatives to zirconium-based and zircaloy cladding so that chemical reactions that generate hydrogen is prevented. We (as an industry) need to accelerate development and deployment of non-hydrogren producing cladding materials; that is, assuming that the coolant/ moderator/ reflector remains (light) water.

2) Having multiple (reactor) units at one site, having more than two units on site needs critical review in terms of post-accident response and management. We must consider the energetic events at one unit exacerbating the situation (safe shutdown) at the other.

3) Further, there is a definite need for a backup (shielded) reactor plant control center that is offsite (remote) so that the accidents can be managed with partial to full extent of reactor plant status (P, T, flowrates, valve status, tank fluid levels, radiation levels).

4) There is a need for standby back-up power, via diesel generator and battery power, at a minimal elevation (100feet/31m) above and some distance from the plant (thus remotely located). This is needed to offset loss of off-site power for plants subject to environmental water ingress (foremost tsunami). Spare battery power should also be kept off-site and in a confirmed ‘charged’ state.

5) It is clear that the spent fuel pool (SFP) cannot be in proximity of the reactor core, reactor pressure vessel or containment itself. The SFP, in current form, is essentially an open volume subcritical assembly that is not subject to design requirements generally defining a reactor core.

Yet, unless thermohydraulic cooling is maintained, it is subject to the similar consequences as a reactor core without adequate cooling. Therefore, we need new passive designs of the SFP, away from the actual plant’s reactor core.

6) Thus needs to be a re-definition of the spent fuel pool. A new standard and design requirement is needed for the spent fuel pool. It should be ‘reclassified’ as a subcritical assembly with a potential to go critical with no active or passive control (rod or soluble ‘poison’) mechanism. Further it needs to be some distance from the reactor plant.

7) We need to identify key valves for emergency core cooling and require them to be non-electrically activated. Otherwise these valves need a secondary means of open and closed status that is remotely located.

8) If an ‘in-containment’ SFP is maintained, then the fuel transfer crane system must be designed so that it is available to remove the fuel during a post-accident phase. OR a second means such as a robotic arm needs to be available.

9) There needs to be a volumetric guidance analysis for ultimate (decay heat) cooling contingency plans so that not only limitations on volume are understood but also transfer of liquids from one volume to another.

Spare tanks and water-filled tanks need to be kept on site as uptake tanks for ‘runoff’ in case of addition of cooling during accident management phases. Spare means to produce boric acid needs to be available off-site. Earthquake-proof diesel generator housing also need to be water-proof. Remote diesel generators are also needed with access to equally remote diesel fuel tanks (also see 4).

10) For nuclear power plants located in or near earthquake zones, we cannot expect structural volumes and ‘channels’ to maintain structural integrity. We should also expect the immediate ground underneath these structures to be porous (earth). Thus design of these volumes and channels should be such that they minimize connections to other (adjacent) volumes from which contaminated (liquid) effluents can flow.

11) Color-code major components so that in case of an accident such as the Fukushima NPP accident, we will be able to quickly identify the major components from digital images.

12) An international alliance of nuclear reactor accident first responders and thereafter, a crisis management team is needed. This does not seem to be available at any significant level at this time. We (the global nuclear industry) cannot wait 3 weeks for international participation.

13) We should consider and work toward international agreement on standards for regulated levels of radiation (activity) and radiation exposure to the general public and separately, those under emergency and extended ‘recovery’ phases.

We should also be consistent in definition and practice of evacuation zoning. We should also strongly encourage acceptance and use of SI unit for activity and exposure and not use culturally-based numbering customs (in Japan, one counts in orders of (‘man’)104, (‘oku’)108, 1012 etc.)

14) Under emergency and crisis management, wider access roads are needed to and from NPPs. The access roads need to be clear of debris and of such width to accommodate large-scale trucks needed as first response and thereafter. A means to access the plant via water, such as ocean, calls for infrastructure (boats, water-containing barge, jet-skis etc) is needed as part of a contingency plan for those plants located near bodies of water.


* Akira T. TokuhiroAuthor ID: Akira T. Tokuhiro (right) (email: (web site) Department of  Mechanical Engineering, University of Idaho, 1776 Science Center Drive, Idaho Falls, Idaho 83402 USA

Keywords: nuclear power plant, accident, meltdown, spent fuel pool, loss of off-site power, earthquake, tsunami

Submitted as short communication to: Nuclear Exchange, First published at April 2011. Reprinted in electronic form at Idaho Samizdat with permission of the author and publisher.

# # #


crf said...

Something about delivery of offsite power, through power cables, should be mentioned.

One of the first lines of defence that was breached was available offsite power, because of downed power lines I guess.

Could there have been an alternate, perhaps underground, grid connections to the nearest power substations?

Alan said...

Great post. I'm glad to see another lessons learned post, and I'm working to compile a few of these lists and make a post of my own. It's taken a long time to get some of these details together, and some of them are still not clear to me. Regarding those mentioned in this post:

8 - What exactly is in mind for a fuel transfer crane that won't be damaged in the event of something like a Hydrogen explosion? I don't understand what the alternatives, and I don't know how you would design it so a crane can be brought in externally.

10 - If structures can't be expected to withstand the earthquake, then that sounds like a problem all-around. What are the flow paths/channels that are of concern? I struggled with the wording of this point. Is the concern water leaking into the ground? Is this with the difficulty of dewatering activities in mind? I don't know what you would do to help that. If you don't have a flow channel, what do you have?

Point 11 seems strange to me, because in my experience they already color-code everything. I know this is true for at least some plants in both the U.S. and Japan, but of course I can't speak for all of them.

Also, regarding point 1: Is this to say that Hydrogen management alone isn't good enough? One option is to change the cladding type, yes, but what about the option of installing more passive valves and assured secondary containment venting? Obviously changing the fuel is a huge deal, I'm interested to know if people believe that the issue can be solved with Hydrogen management, or if full elimination is considered necessary.


Joffan said...

These are excellent discussion points, but I do not think that all of them would qualify - yet, at any rate - as "lessons learned".

In particular there is the need to balance risk in normal operation against risk in emergency situations, where emergency provisions introduce difficulties or new risks into normal operations.

From my layman's perspective, I woult respond to the points raised as follows:
1) Alternative claddings have been and are being researched; I don't know how close they are to common availability.
2) Balance multi-unit sites against multiple sites
3) Remote control center relies on communications and sensors; not always usable as at Fukushima.
4) Remote standby power; possibly. Salt-water immersion capability might be another route. Floatable facilities another.
5) Spent fuel pool location; definitely a case to be considered, but with regard also to normal operations risks.
6) Rethinking spent fuel storage; yes - also faster more routine use of dry cask storage. Rethink schedule for air cooling.
7) Non-electrical valves; not sure. Worth a look but valve operation really should not need huge power. Hydraulic backup?
8) Fukushima experience argues that this point is of limited value
9) I don't really understand the volumetric guidance point.
10) This is perhaps a more extreme position on ground structures than is sensible under risk-based assessment. In fact at Fukushima it may have been an advantage that these volumes were connected.
11) Not sure color coding is the best way to achieve this. RFID?
12) First responders: Good talking point for IAEA, but site personnel would normally need to lead actions. International rescue robots? :-)
13) No argument with working towards rational levels of radiation exposure.
14) Roads/access used for construction should normally be adequate for subsequent plant access. Road clearance certainly needs to be in emergency plans.

netudiant said...

This is a very helpful overview.
However, the list presents a daunting series of tasks, which in aggregate represent a restart for the industry.
The ticket for this list is huge, even if it is simply used to guide the replacement reactor program.
It is so big that imo it might revitalize the search for alternative reactor designs, whether CANDU, thorium molten salt, pebble bed or whatever, because the LWR with all these provisions will be an expensive facility.

Marcel F. Williams said...

Nuclear reactors need to be built small enough so that when the power plant shuts down, the reactors can cool down naturally without the aid of any external power source or human intervention.

Building such small reactors underground also protects them from storms, floods, tsunamis, airplane crashes, and potential terrorist attacks.

Anonymous said...

1. The technical lessons in the post are only one part and not necessarily the only part that can be improved. Sufficient international audits as they are practiced in other areas, such as information security and finance could help to assert details of operational status and uncover gaps and necessary improvements before an accident occurs.

2. The late use of robots in Fukushima is tough to understand specifically for the reconnaissance aspect. The tasks of reconnaissance similar robots perform in the middle east are not that different.
Appropriate shielding of their electronics as needed could most likely be addressed in less than 3 weeks, no?

Dr. Martin Lades